A comparative study on the construction of age estimation regression models based on the pulp dentinal index of the first permanent molar
10.13618/j.issn.1001-5728.2025.03.007
- VernacularTitle:基于第一恒磨牙髓室牙本质指数构建年龄推断回归模型的比较研究
- Author:
Yanjie DING
1
;
Tao ZHONG
1
;
Wei WANG
1
;
Yuxin HE
1
;
Xiao ZHANG
1
;
Shilin ZHANG
1
;
Wenli SHI
1
;
Bo JIN
1
Author Information
1. 川北医学院基础医学与法医学研究所,四川南充 637000
- Publication Type:Journal Article
- Keywords:
Forensic anthropology;
Age estimation;
CBCT;
First permanent molar;
Pulp dentinal index
- From:
Chinese Journal of Forensic Medicine
2025;40(3):302-307
- CountryChina
- Language:Chinese
-
Abstract:
Objective To compare regression models for age estimation constructed based on the Pulp Dentinal Index(PDI)of the first permanent molars and explore a more accurate and applicable novel method.Methods A total of 900 Cone Beam Computed Tomography(CBCT)image datasets from adult Han Chinese individuals(455 males and 445 females)residing in Sichuan province,China,were collected from the Affiliated Hospital of North Sichuan Medical College.The PDI of the first permanent molars were measured using Mimics software.Regression analysis was performed with age as the dependent variable and PDI as the independent variable to establish and validate the optimal mathematical model for age estimation.Results Strong correlations were observed between the PDI of all four first permanent molars and age,with the correlation being stronger for maxillary teeth compared to mandibular teeth,and the highest correlation found in the left maxillary first molar among females(r=0.881).Significant differences in PDI were identified between maxillary and mandibular positions as well as between genders(P＜0.05),but not between left and right positions(P＞0.05).Among the 11 mathematical models constructed,the cubic regression model outperformed others,with the left maxillary first molar model demonstrating the best performance(Age=73.93-70.79x-68.75x2+94.33x3),yielding a mean absolute error(MAE)of 4.88 years.Conclusion Among the 11 regression models constructed in this study based on CBCT-measured PDI values of the first permanent molars,the cubic regression model exhibited the highest accuracy,with an MAE of 4.88 years.
